Noninvasive assessment of central and peripheral arterial pressure (waveforms): implications of calibration methods.

OBJECTIVES: Noninvasive estimation of central blood pressure (BP) from radial artery pressure waveforms is increasingly applied. We investigated the impact of radial artery waveform calibration on central BP assessment and calculated pressure amplification, with focus on the one-third rule used to estimate mean arterial BP (MAP). METHODS: Pressure waveforms were noninvasively measured at the radial and carotid arteries in 1873 individuals (age 45.8+/-6.1 years). Radial and carotid artery waveforms were calibrated using brachial artery DBP and SBP, MAP estimated with the one-third rule and MAP estimated as brachial DBP along with 40% of brachial artery pulse pressure. RESULTS: Central SBP obtained via a transfer function was 123.5 +/- 15.7, 117.8 +/- 14.2 and 126.0 +/- 15.4 mmHg (mean +/- SD) following above-mentioned three calibration schemes, respectively. Using the same calibration schemes, carotid artery SBP was 131.4 +/- 15.2, 118.4 +/- 14.4 and 126.8 +/- 15.7 mmHg, respectively. Central-to-brachial amplification was 13.0 +/- 3.6 mmHg using second method as compared with 4.6 +/- 3.8 mmHg with third method. Brachial-to-radial amplification was actually negative (-6.3 +/- 4.5 mmHg) using second method, whereas 3.4 +/- 5.5 mmHg was found with third method. CONCLUSION: Both carotid artery SBP and central SBP obtained via a transfer function are highly sensitive to the calibration of the respective carotid artery and radial artery pressure waveforms. Our data suggest that the one-third rule to calculate MAP from brachial cuff BP should be avoided, especially when used to calibrate radial artery pressure waveforms for subsequent application of a pressure transfer function. Until more precise estimation methods become available, it is advisable to use 40% of brachial pulse pressure instead of 33% to assess MAP.

Distance measurements for the assessment of carotid to femoral pulse wave velocity.

BACKGROUND: Carotid-femoral pulse wave velocity can be determined using different distances - either direct carotid-femoral distance or subtracted [(sternal-femoral) - (carotid-sternal)] distance - resulting in pulse wave velocity differences of up to 30%. The present study aims to present and validate a population-based model for the conversion between distances. METHOD: Three thousand one hundred and sixteen participants from the Asklepios study (n = 2510) and Hôpital Européen Georges Pompidou (n = 606) databases, in which all distance measurements were available, were randomly distributed in a model (n = 311) and validation (n = 2805) population. Model parameters for the conversion equations were selected and evaluated using multiple linear regression with stepwise selection of covariates (age, sex, weight, height, BMI and waist circumference). The proposed model was evaluated on the validation population. RESULTS: The difference between direct and subtracted distances was found to be partially dependent on body height, and its inclusion in the multivariate model improved model performance by over 20%. Other combinations of adjustments did not improve model prediction. Conversion equations derived in the model population were: Estimated Direct_distance = 0.45*Subtracted_distance + 0.21*height + 0.08 and Estimated Subtracted_distance = 1.04*Direct_distance - 0.11*height - 0.02, respectively. Applying these equations for estimation of direct and subtracted distances in the validation population yielded good correspondence to measured results (r = 0.73 and 0.57, respectively), with nonsignificant mean differences between estimated and measured values. Increasing the size of the model population did not significantly change the model validity. CONCLUSION: In cases in which not all distance measurements are available for exact conversion, the presented equations can be used to convert between distance definitions.

Assessment of pressure wave reflection: getting the timing right!

Assessment of timing and magnitude of wave reflection is ideally based on wave separation analysis (WSA). In clinical practice, however, waveform analysis (WFA) is often used to study wave reflection, with different coexisting approaches to assess 'landmarks' on the waveform which are indicative for return of the reflected wave. The aim of this work was to compare WSA and WFA. Data were obtained from 2132 subjects (1093 women) aged between 35 and 56 and free from overt cardiovascular disease. Carotid pressure and aortic flow waveforms, and carotid-femoral pulse wave velocity were measured non-invasively. WSA yielded the timing of return of reflected wave (T(f-b)), the ratio of forward and backward pressure wave (P(b)/P(f)), and the effective length of the arterial tree (L(eff)). WFA resulted in identification of the shoulder (T(sho)) or inflection point (T(inf)) as landmark points, with subsequently derived augmentation index and L(eff) (AIx(sho) and L(eff,sho), AIx(inf) and L(eff,inf), respectively). (i) Neither T(inf) nor T(sho) corresponded with the timing obtained from WSA. (ii) Measurements of L(eff) were found to decrease with age (conforming with current physiological insights) whilst L(eff,inf) was found to increase with age in women, and mixed results were obtained for L(eff,sho). (iii) Both AIx(inf) and AIx(sho) showed a persistent gender difference which was not present in P(b)/P(f). Using the pressure at T(f-b) to calculate AIx, the systematic gender difference in AIx(f-b) was greatly reduced. Analysis of pressure wave reflection is optimally based on measurement of pressure and flow, rather than on waveform analysis alone.